Opinion. Isolated major congenital heart disease

Transcription

1 Ultrasound Obstet Gynecol 2001; 17: Opinion Blackwell Science, Ltd Isolated major congenital heart disease The principal theme of this issue of the Journal is the prenatal detection of fetal cardiac defects which remains the most problematic aspect of prenatal diagnosis. It is well recognized that cardiac malformations always come lowest in terms of prenatal detection. However, the situation is compounded by the fact that at least 50% of cardiac malformations are minor and have little long-term consequences for the developing child, so realistically, they should be excluded from discussion of detection rates. Furthermore, many major cardiac malformations are associated with chromosome abnormalities, genetic syndromes or are part of a multiple malformation disorder. The detection of cardiac malformations in these circumstances is usually dependent upon the recognition of the associated problems. Finally, of course, there is the question of whether there are good antenatal predictive criteria at 20 weeks gestation on the postnatal outcome of major congenital heart disease (MCHD) to justify counseling couples for continuation of the pregnancy. Several papers in this Journal address some of these contentious issues. To complement these papers, I felt it would be worthwhile to ask some of the luminaries in the field of ultrasound detection of fetal cardiac malformations to answer some of the key questions on this topic, which are still the subject of controversy. I have confined the questions to the prenatal detection of isolated MCHD. Their answers are summarized here; full answers will be given on the Internet at the following site: UOG439sm.htm. Although divergence of opinion on many aspects of policy for screening and diagnosis was expected, I was surprised that there was so little agreement on such fundamental issues as the definition and prevalence of MCHD and the detection rates for this condition. I believe that the time has come for ISUOG to set up a fetal cardiology working party to try and achieve a consensus view on some of these current areas of disagreement, without which we will fail to make progress in the detection and management of isolated major congenital heart disease. I would like to thank Lindsey Allan (LA), Beryl Benacerraf (BB), Joshua A Copel (JAC), Julene S Carvalho (JSC), Rabih Chaoui (RC), Sturla H Eik-Nes (SE-N), Eva Tegnander (ET), Ulrich Gembruch (UG), James Huhta (JH), Gianluigi Pilu (GP), Juriy Wladimiroff (JW) and Simcha Yagel (SY) for their prompt and thoughtful responses to the questions. S. Campbell Editor-in-Chief Question 1: Is there a definition of major congenital heart disease (MCHD) and if so, what is the birth prevalence of isolated MCHD? LA: There is no universally accepted definition of MCHD. It has been variously defined as: 1 CHD resulting in presentation to a cardiologist during infancy; 2 CHD requiring intervention during the first 6 months; 3 CHD requiring intervention during the 1st year of life; 4 an abnormality of connection. None is a perfect definition of what most cardiologists understand as MCHD which is really CHD which will have a significant effect on a child s life although they all come close. The first three are a little too broad and could include some non-major lesions such as a ventricular septal defect (VSD) and the last is a little too narrow, as, for example, critical aortic or pulmonary stenosis would not be connection anomalies but are certainly severe and life-threatening. The birth prevalence of MCHD is about 3/1000 live births but a significant proportion of those will not be truly isolated but will be associated with other structural malformations or chromosome anomalies. As chromosome analysis becomes more refined, the incidence of associated chromosomal anomalies is increasing, particularly in relation to the incidence of microdeletion of chromosome 22 with some relatively common heart conditions. BB: The definition of major congenital heart disease in my opinion is a heart defect that has the potential to be lifethreatening and requires medical and/or surgical intervention after birth. There are many reports in the literature concerning the estimated prevalence of major congenital heart disease in the newborn population. Obviously it is considerably higher in the fetal population due to attrition of fetuses before they become newborns. Most reports of live births with major heart defects indicate a prevalence between 5/1000 and 9/1000. JAC: I don t know a formal definition, but clinically tell parents that anything I can see prenatally is major. I would exclude small VSDs from the major group. Birth prevalence surveys, which should capture mostly major disease, suggest that the prevalence of CHD is about 5/1000 in the US and Europe. JSC: I do not believe that there is such a definition. Perhaps there is a consensus view being formed over the years which includes as MCHD those cases that will require surgical or catheter intervention usually (but not necessarily) in the first year of life. To these, I would also add certain conditions which may be well balanced and are less likely to require 370 OPINION

2 intervention but need long-term follow up (e.g. corrected transposition without major associated abnormalities). Birth prevalence for all cases of CHD is generally accepted to be around 8/1000 live births. It is reasonable to assume that MCHD contributes to at least 4/1000. RC: Major CHDs are defects which manifest clinically in the first days and weeks of life, leading to a referral to a specialized center. Furthermore these are the defects which need cardiac surgery. This includes thus all defects with a need for palliative surgery (e.g. Norwood, Fontan) and some defects undergoing corrective surgery. The Baltimore Washington Infant Study 1 is considered as the most reliable and largest study on the birth prevalence of children with cardiac defects ( ). According to this study the following heart defects mostly considered as minor have the following rates among liveborns with a CHD: VSD 32%, pulmonic stenosis 9%, secundum atrial septal defect 7.7%, patent arterial duct 2.4%, bicuspid aortic valve 1.9%. If one adds to this group of 53% some children with mild aortic stenosis (2.9%), mild aortic coarctation (4.6%) and mild forms of atrioventricular septal defects (7.4%) and others the birth prevalence of the remaining major congenital heart defects can be estimated to be 40% of all CHD. In the study the birth prevalence of CHD was found to be 4.9/1000 livebirths. Major CHD has thus a prevalence of 2 out of 1000 livebirths. SE-N + ET: We have a problem with the expression birth prevalence in the above question. It makes it difficult to comment since most modern and reliable data include prenatally diagnosed conditions and thus not necessarily born individuals. In our practice we use the definition of Mitchell and Korones 2. CHDs are classified as critical or non-critical (major or minor). A defect is classified as critical when a surgical repair most likely is required because of gross structural complexity and functional significance, e.g. transposition of the great arteries, hypoplastic left heart syndrome, atrioventricular septal defect, coarctation of the aorta, large VSD. A CHD is classified as non-critical when no intervention is likely to be required, e.g. mild pulmonary stenosis, mild aortic stenosis, small VSD, small atrial septal defect. If there is doubt as to whether a heart defect is critical or non-critical, the defect is classified as critical if intervention is actually needed during the first month of life. In a recent well-controlled population of patients screened antenatally, we had 40 MCHDs which gives an incidence of 3.3/1000. There were 27 isolated MCHDs (equivalent to 68% of the population of MCHD) which gives an incidence of 2.2/ UG: Unfortunately, there is no commonly accepted definition of MCHD although the inclusion and the exclusion of several cases of minor cardiac defects have a considerable impact on study results about the validity of prenatal ultrasound in general and of cardiac ultrasound in particular. The categorization of minor and major congenital heart disease is usually based on the severity of the cardiac defect, especially on the necessity for operative and/or interventional treatment to enable further survival. Following this classification small ventricular septal defects, mild stenoses of the semilunar valves, and mild aortic coarctation, but also mild forms of Ebstein s anomaly are minor or better mild cardiac defects, whereas the major congenital heart defects include all defects with veno-atrial, atrioventricular, or ventriculoarterial discordance, large ventricular septal defects, and hemodynamically relevant anomalies in the region of the aortic arch. However, this important differentiation into major and minor congenital heart defects should not be applied uncritically to the assessment of efficacy of cardiac screening and detailed echocardiography during fetal life because some minor defects, such as primum atrial septal defects and small ventricular outflow tract defects, may be associated with chromosomal and non-chromosomal extracardiac anomalies and therefore prenatal detection is highly relevant for good prenatal and perinatal care. The incidence of MCHD is approximately 3 4 per 1000 live births and 5 per 1000 fetuses in the second trimester. JH: Major CHD is that which will require cardiac surgery in the first year after birth. It is approximately one half of all CHD or about 4 5 per 1000 live births. GP: The term major is conventionally used to indicate an anomaly with surgical implications. In reality, with current developments in pediatric cardiology and cardiac surgery many cardiac malformations have either minimal or no impact at all on the duration and quality of life. While dealing with fetal diagnosis, it would be reasonable to limit the term major to the lesions that continue to be associated with significant mortality or morbidity. Congenital heart disease occurs in about 5 cases out of 1000 births, 25 30% of cardiac malformations are associated with extracardiac malformations, and about 50% of cardiac malformations are not life-threatening with current treatment 1. Therefore, isolated major cardiac anomalies occur roughly in 2 cases out of 1000 births. The incidence is likely to be slightly greater in fetuses at midgestation because some cardiac anomalies are associated with intrauterine demise. JW: Major congenital heart disease can be defined as heart disease which needs perinatal treatment (mostly surgery). Whether perinatal surgery followed by a normal life expectancy must also be categorized as MCHD is subject to debate. I would estimate the prevalence of MCHD to be 4 per 1000 live births. SY: I would define as MCHD, those lesions which will require surgical intervention in the first year, or cardiac disease that will affect the child s quality of life, such as cardiomyopathy. The birth prevalence of such (isolated) defects is approximately 2 : 1000 live births. Question 2: What prenatally diagnosed MCHDs may have their outcome improved by altering perinatal management? LA: All forms of CHD which are dependent on the patency of the arterial duct in the immediate postnatal period or are mixing lesions such as transposition of the great arteries or total anomalous pulmonary venous drainage may have an improved outcome by altering perinatal management. Those lesions which are duct-dependent include any form of aortic or pulmonary stenosis or atresia and coarctation or interruption Ultrasound in Obstetrics and Gynecology 371

3 of the aorta. Hemodynamic status has been shown to be better in all those types of heart disease when prenatally diagnosed, and this is likely to translate into improved long-term morbidity. Improved mortality has been clearly shown as a result of prenatal diagnosis in the outcome of transposition of the great arteries. BB: There is some controversy regarding the potential improvement in outcome of fetuses who have their heart disease diagnosed prenatally. I firmly believe that there are some specific heart lesions such as the ductus-dependent defects, where prenatal diagnosis can have a major impact on the outcome of the newborn. Knowledge of a ductus-dependent lesion such as tetralogy of Fallot, transposition of the great arteries, or coarctation of the aorta allows the delivery to be done in a center where cardiology services are available or immediate transport of the child can take place. The ductus can be kept open until the work-up is performed so that the child is not increasingly sick and acidotic by the time the diagnosis is made. JAC: Our data from Yale 4 showed significantly improved survival for lesions amenable to biventricular repair, but we did not have statistical power to distinguish among specific lesions. Based on clinical observations, and the data from France 5, however, I believe that the single most important lesion to diagnose prenatally is transposition of the great arteries with an intact ventricular septum. JSC: The one condition that has been shown to have its outcome improved following fetal diagnosis is transposition of the great arteries some cases may need urgent balloon atrial septostomy soon after birth. Unfortunately, transposition is not usually detected prenatally. One would also have expected that optimization of perinatal management with the elective use of prostaglandin would have improved results of surgery in duct-dependent lesions. This is not necessarily the case. In hypoplastic left heart syndrome for example, some groups have shown worse outlook for cases prenatally diagnosed while others have shown improved outcome. RC: Most complex cardiac defects can have an improved outcome, when delivered at centers with pediatric cardiology. Especially cardiac anomalies with ductus arteriosus dependency, with clinical deterioration in the first hours of life profit from an elective treatment before critical symptoms occur. In a study of Copel et al. 4 it was shown that only for heart anomalies undergoing biventricular repair did prenatal diagnosis have an improved outcome statistically. Some other studies emerged reporting on transposition of the great arteries and other anomalies as well. SE-N + ET: We believe the following forms of MCHD have their outcome improved following prenatal diagnosis: 1 transposition of the great arteries; 2 CHD with ductus-dependent pulmonary circulation: pulmonary atresia with intact ventricular septum or restrictive VSD; tricuspid atresia with restriction at the level of the atrium or ventricle; 3 CHD with ductus-dependent systemic circulation: hypoplastic left heart syndrome; interrupted aortic arch. UG: Termination of pregnancy secondary to prenatal diagnosis of isolated and non-isolated cardiac defects may considerably influence the data on perinatal outcome in newborns with cardiac defects in general and in specific lesions in particular. Regarding only MCHDs, however, it should be reasonable to assume that an optimized perinatal management strategy secondary to an adequate prenatal diagnosis, i.e. helping to avoid the hazards of neonatal transfer and clinical worsening with the possibility of hypoxia and acidosis causing multiorgan failure and long-term neurological damage, significantly improves the quantity and quality of short- and also longterm survival especially for duct-dependent cardiac defects, i.e. severe obstructions of the right and left ventricular outflow tracts and of the aortic arch, and for the transposition of the great arteries and total anomalous pulmonary venous return. Analyzing the sparse data in the literature, an improvement regarding mortality and short-term morbidity seems to be confirmed for dextro-transposition of the great arteries, partly for critical ventricular outflow tract obstructions and also for hypoplastic left heart syndrome. JH: The data so far show significant improvement in the preoperative condition of neonates with transposition of the great arteries, hypoplastic left heart syndrome, ductaldependent pulmonary atresia, and total anomalous pulmonary venous connection. Fetuses with heterotaxy can all have their management improved by accurate prenatal diagnosis. GP: Obstetric management is unlikely to influence the outcome of cardiac lesions, with the possible exception of severe dysrhythmias (tachycardias and complete heart block). Fetuses with ductal-dependent malformations may theoretically benefit from prompt referral to a tertiary care center, although thus far no clear cut evidence exists. Of course, the option of termination of pregnancy in the presence of a severe fetal malformation should be regarded as a possible benefit to the couples. JW: Transposition of great arteries; aortic stenosis; coarctation of the aorta; pulmonary stenosis; tetralogy of Fallot and perhaps hypoplastic left heart syndrome. SY: It has been proven that prenatal diagnosis of transposition of the great arteries improves perinatal morbidity and mortality rates, particularly in the case of patients living in rural areas, distant from suitably equipped medical services. Planned delivery in a tertiary center where speedy intervention will be possible without transfer and delay, or preventing discharge from hospital with concomitant worsening of the neonate s preoperative condition, has been shown to improve surgical outcome. In the case of most other lesions, while this may be true, it has yet to be proven. Theoretically, for all lesions requiring surgical repair and/or immediate neonatal intervention (e.g. ductus-dependent lesions) prevention of delay in treatment would improve the chances of successful correction by improving the neonate s overall preoperative state. On the other hand, early identification of affected fetuses allows for timely decision to terminate a pregnancy. In cases of severe structural malformation, lethal defects or those with dismal prognosis for repair or a high incidence of associated developmental delay or significantly affected quality of life, termination is justified. 372 Ultrasound in Obstetrics and Gynecology

4 Question 3: There is a wide variation in the estimated sensitivity of routine MCHD screening. What are the reasons for this? LA: Differing scanning skills, protocols and supervision of standards in general ultrasound screening account for differing success in detecting MCHD, as well as other anomalies. It has been well shown by several authors that cardiac scanning skills can be learnt in a suitable environment, they are not prohibitively time-consuming and excellent levels of detection can be achieved if there is motivation and effort applied to the task. Training programs, however, must be available, enlightened and continually refreshed. BB: Sonographic evaluation of the fetal heart is by far the most complex area of the fetus to master. Accurate evaluation of heart anatomy requires a minimum of expertise, which many practitioners do not achieve unless they become specialists or specifically dedicate time and effort to fetal imaging beyond what is normally required by standard training programs. I strongly believe that the differences in the practitioners ability to identify heart defects is in large part related to the difference in training and skill that is present in our discipline due to the very heterogeneous mix of practitioners that do fetal imaging. It must be kept in mind that most fetuses with isolated congenital heart disease do not have an increased risk for this disorder; therefore, for routine screening to be successful, a comprehensive examination of the heart on all routine scans would be necessary. Unfortunately, this is not currently being done in most practices. JAC: Operator experience is primary. Variations in equipment quality, gestational age at screening and patient body habitus are also factors. JSC: The main reasons are the low case ascertainment in many studies and no uniformity about which conditions should be included or not. If there is low ascertainment the sensitivity will be higher and if minor defects are included, the sensitivity is likely to be lower. RC: Screening methods measuring values (like AFP, glucose-challenge test, etc.) rely on objective measurable values which are subject to quality control. Ultrasound skill and visualization is dependent on many variables beginning with the examination conditions like gestational age, maternal obesity, fetal position, duration of the examination and some technical aspects like the skill and experience of the examiner, the quality of the equipment and the transducer frequency. We cannot as yet standardize these variables to make a reliable comparison of sensitivities between different examiners and centers. One of the major teaching purposes is to train people not only how to examine the fetal heart but to demonstrate pathological appearances. SE-N + ET: There are a number of reasons: Data based on unselected populations give varying results, but in most series, it is difficult to know whether or not the addressed population is unselected. Many series do not sufficiently ascertain the complete number of CHDs in the population. A further problem is that many series do not specify what was diagnosed at the time of the routine echocardiography (around weeks) and what was diagnosed later in pregnancy. Other factors affecting detection rates are the gestational age at the time of the routine scan, the completeness of the cardiac examination (four-chamber view, outlets, etc.) and the definition of MCHD. UG: There are three main reasons for the wide variation in the estimated sensitivity of routine MCHD screening. 1 The screening protocol employed. Without firm integration of the four-chamber view into the sonographic screening, the detection rate of MCHD is approximately 5%. Using only the four-chamber view for cardiac screening the detection rate for MCHD can ideally extend to 40 50% whereas by incorporating the demonstration of both ventricular outflow tracts and the great arteries a maximum of 70 80% of MCHDs can be detected at around the 20th week. If color flow mapping is added, nearly 90% of the MCHDs may be detectable at around the 20th week. 2 There is a wide variation in the general level of training of the participating examiners between and even within the reported screening studies for MCHDs. 3 The examination periods have varied in the reported studies and for cardiac screening the gestational age has a major influence on the results. The best time period seems to be between 22 and 24 weeks gestation; firstly because of the in utero development of secondary structural anomalies, i.e. hypoplasia of the ventricle and/or great artery, hypertrophy of the ventricular wall, atrial dilation and cardiomegaly, and also because of the larger size of the cardiac structures, the detection rate is higher in this time period. Further reasons for the variation in the detection rates in the reported studies are: different definition of CHD and MCHD; retrospective vs. prospective study design; timing and quality of postnatal examinations and also of the quality of the autopsy examination of abortions and stillborns. JH: Untrained, inexperienced ultrasonographers and poor training by the cardiologist. GP: Antenatal MCHD screening is almost universally based upon the four-chamber view of the fetal heart. Surprisingly enough, there is no clear agreement on the definition of a proper four-chamber view. In some settings, sonographers are satisfied with a cursory inspection of the beating heart roughly demonstrating that it is made up of four chambers. In other settings, the position of the heart inside the chest is identified and matched with the topography of the abdominal organs, the cardiac axis is estimated, the size of the chambers is estimated, cine-loop is used to demonstrate the opening movement of the atrioventricular valves and the integrity of the septa and possibly color Doppler is also used to demonstrate atrioventricular flows. Of course the expertise of the operators is a critical factor, but the lack of standardization certainly contributes to the wide discrepancy in detection rates of different studies and the disappointing results of many multicenter studies. JW: The wide variation in the estimated sensitivity of routine MCHD screening is determined by several factors, such as scanning experience, standard or ultrasound equipment, number of cardiac views (four-chamber view including inand outflow tract vs. four-chamber view only), quality and duration of cardiac ascertainment after delivery. Ultrasound in Obstetrics and Gynecology 373

5 SY: The widely varying levels of training and continuing medical education support available to examiners around the world have led to equally widely varying levels of expertise. In addition, varying demands of different medical services delivery and referral systems expose very different populations to fetal echocardiography screening. Question 4: Do you recommend that all pregnant women should be routinely screened for MCHD? Or should targeted screening be employed, and if so what are the recommended indicators of risk? LA: All pregnant women should be routinely screened for MCHD. BB: I have advocated for many years that all fetal sonograms at 18 weeks or more (even earlier when possible) include four-chamber and both outflow tract images. Using these three standard views on all pregnant women, detection of more than 80% of major fetal heart defects can be achieved even on low-risk patients. Targeted screening can then be employed for patients who have an abnormal fetal survey and require further evaluation. This more in-depth or targeted screening of the heart would include the use of color Doppler, pulsed Doppler, M-mode, outflow tract measurements, etc. JAC: I d like all pregnant women to have a high quality anatomic survey. Due to time and training issues though I still think we must reserve full fetal echocardiograms for standard maternal and fetal risk factors (e.g. family history, drug exposure, metabolic abnormalities, not just diabetes but now women who have mutations of the methylenetetrahyrofolate reductase gene). JSC: In societies where there are screening programs, all women should be offered routine screening. Currently, specialized fetal echocardiography is usually restricted to recognized high-risk groups. The way forward to increase detection of MCHDs is to improve the effectiveness of screening programs so that a higher number of cases from low-risk populations is referred for a specialized scan. For isolated MCHDs, the main indicators of risk remain family history, particularly if the mother or two siblings are affected (up to 10%) and increased nuchal translucency (NT) with normal karyotype (at least 5% if NT > 3.5 mm). RC: In countries where screening ultrasound in pregnancy is established there is no excuse not to use it for examination of the fetal heart in all pregnancies. The positive aspect of screening for cardiac defects compared with other anomalies (e.g. Down syndrome, neural tube defects) is that it does not automatically result in termination of pregnancy in most cases. It is one of the anomalies where optimizing management of the neonate in the perinatal period could improve outcome. Targeted examination of groups at high risk should in addition be performed at centers with more experience with fetal echocardiography. The yield (incidence) is lower than 5% in most classic situations like positive family history, diabetes, arrhythmia and drug intake. These are mainly done to reassure the pregnant women. More effective indications with a higher incidence of congenital heart defects are referred fetuses detected at screening with extracardiac anomalies, non-immune hydrops, suspected abnormal heart anatomy and significant increase in nuchal translucency thickness in early pregnancy. SE-N + ET: We recommend that all pregnant women are routinely screened for MCHD. Targeted screening has proved to be of low value due to the fact that most CHD is found in pregnancies at low risk. UG: I recommend general screening including the fourchamber view and the outflow tract views for all pregnant women at weeks gestation because a targeted echocardiography in high-risk pregnancies only (increased risk based on family history, maternal diseases, exposure to teratogens in pregnancy, documented fetal anomalies more frequently associated with cardiac defects) will detect approximately 10 15% of the cases with CHD. Without routine screening the vast majority of fetuses with isolated MCHDs which could benefit from optimized perinatal management would not be diagnosed prenatally. Therefore, an extended cardiac examination should become a firm part of second-trimester fetal malformation screening. JH: Yes all women should be screened. Targeted screening for reassurance should be done with family history of CHD, elevated Hemoglobin A1C, family history of cardiomyopathy, sudden death, long Q T syndrome, or other congenital malformation. GP: In those countries where universal sonographic screening for fetal malformations is accepted, it would be unreasonable not to look for cardiac malformations. JW: I would argue against routine screening of MCHD because of the complexity of fetal echocardiography and therefore low pick-up rate due to the low prevalence of the condition. In targeted screening, the recommended indicators of risk are: increased nuchal translucency, juvenile diabetes mellitus, previously affected infant, maternal/paternal congenital heart disease, certain medications (antiepileptic drugs, lithium, alcohol), collagen diseases (LE), drugs (heroin, etc.), viral infections, fetal cardiac arrhythmia, fetal hydrops, chromosomal anomaly, extracardiac structural anomalies. SY: I heartily recommend that all gravidae be screened for all structural birth defects amenable to prenatal diagnosis, not only CHD or major CHD, early enough in pregnancy to allow parents the option of karyotyping affected fetuses, and elective termination where indicated. Even if we limit our discussion to isolated major CHDs, their prevalence approximates that of Down syndrome. If we consider the enormous resources invested to prevent the birth of Down syndrome infants, it is certainly justified to screen all women for MCHD. Furthermore, it must be remembered that 90% of congenital heart defects occur in low-risk mothers. Question 5: What is the optimal gestational age for screening? LA: At 18 weeks all the cardiac connections can be seen in almost all patients with current equipment in skilled hands. Although image quality (and therefore accuracy and confidence in diagnosis) gets better as gestation advances, the option for pregnancy interruption becomes much more difficult if a malformation is found after 20 weeks. For this reason, I think 18 weeks is ideal even if it means the operator has to try a little harder. 374 Ultrasound in Obstetrics and Gynecology

6 BB: I believe that the optimal gestational age for sonographic screening of the fetal heart is weeks, when the heart is large enough to be seen in detail in the vast majority of patients. Although there are some that advocate a full exam of the heart at 14 weeks, I have seen a few occasions where the four chambers and outflow tracts appeared normal early in gestation; however, changes in the rate of growth of cardiac structures can take place during the remainder of the pregnancy such as aortic stenosis leading to critical aortic obstruction and resulting in hypoplastic left heart syndrome. Differential growth of the two ventricles may change over the course of the pregnancy and this is not always detectable as early as 14 weeks. If the heart is evaluated at 14 weeks, it must be re-evaluated around 20 weeks. JAC: That depends on how we are going to screen. If we are to use nuchal translucency then weeks, though I believe that data require reconfirmation in other populations and countries. If transabdominal sonography then it can be done at weeks, but sensitivity will be better at weeks. JSC: Taking into account that not all patients are easy to scan, optimal routine obstetric screening is best performed at weeks of gestation. RC: Considering the options to be offered to the patient in case of detecting an anomaly (i.e. termination, karyotyping) the best time is still between 20 and 22 weeks. The best visualization is however, achieved between 26 and 29 weeks. SE-N + ET: In a routine setting the time interval weeks is appropriate. Some would prefer 22 weeks for a specific cardiac scan. If only one routine scan is offered in the pregnancy, 22 weeks is too late for the total fetal examination. The early (11 14 weeks) scan most likely is of additional value, but there is little reason to believe it ever can replace a later scan when it comes to the detection of cardiac and any other disorders. UG: The best time period seems to be between 20 and 22 weeks gestation. At this period screening for cardiac defects can become a firm part of a general ultrasonic malformation screening which is already performed in many countries. Because of the in utero development of secondary structural anomalies, i.e. hypoplasia of the ventricle and/or great artery, hypertrophy of the ventricular wall, atrial dilation, cardiomegaly, and also of the larger size of cardiac structures the detection rate is higher at this period than between 16 and 19 weeks and also between 14 and 15 weeks, which is the best period for early echocardiography. Second-trimester screening is important when improvement of outcome by optimizing perinatal management is the reason for prenatal ultrasound investigations. On the other hand, early ultrasound examination including extended echocardiography between 13 and 15 weeks gestation appears to be the preferential strategy for the diagnosis of serious anomalies at an early gestation when termination of an abnormal fetus is more acceptable. For this concept, however, highly experienced examiners, improved technology, and often transvaginal scanning are needed to achieve acceptable results with the reported detection rate between 60% and 70% for congenital heart disease. But early fetal echocardiography should always be performed in selected high-risk cases (increased risk based on family history, maternal diseases, exposure to teratogens in pregnancy; documented fetal anomalies more frequently associated with cardiac defects) including fetuses with increased nuchal translucency and/or early hydrops. JH: weeks. GP: weeks. With MHz abdominal transducers. JW: The optimal gestational age for screening still seems to be weeks of gestation. This is determined by the fetus/amniotic fluid volume relation and the upper limit of 24 weeks for pregnancy termination in many countries. SY: If only one screening exam is to be performed, I recommend no later than 23 gestational weeks, to allow fetal karyotyping and/or pregnancy termination where indicated if MCHD is diagnosed. If two exams will be performed, I recommend early in the second trimester (weeks 14 16), and a repeat exam at weeks. The earlier examination provides a better time frame for both karyotyping and termination when these are indicated, which in the latter instance eases both the physical and emotional difficulty of the procedure and reduces its cost. Early transvaginal scanning cannot be used alone to screen for cardiac defects, because of the developmental nature of many lesions on the one hand, and technical limitations on the other. Question 6: How many cardiac views should be obtained at the screening examination? LA: Cardiac scanning is a continuous process and should not be thought of in views. A sweep from the abdomen to the fourchamber view and on up the chest to display the arterial relations and connections and the duct and arch, and then the same sweep in color, show everything necessary in a normal study. BB: We recommend at least three cardiac views for each screening examination; these include the four-chamber view and both outflow tracts. When possible, we also recommend the aortic arch and ductus arches, although these are not a necessary part of routine screening in our laboratory as are the three abovementioned views. JAC: Ideally I d prefer a four-chamber view and outflows, though standard in the US remains a four-chamber view. JSC: Ideally, the screening would involve real-time assessment with an ultrasound sweep from the four-chamber to the great arteries. In practice, it is the quality rather than the number of views that matters. A departmental policy to include great vessel views in a screening program is unlikely to be successful unless those performing the scan are confident on the assessment of the four-chamber view. A screening program that relies on good standard four-chamber real-time assessment rather than a still frame will certainly improve the current low detection rates. RC: Many cardiac defects are not detectable in the fourchamber view but only by visualizing both great vessels. According to our experience a transverse view of the upper thorax including the three vessels (three-vessel view) should be added to the four-chamber view in screening. SE-N + ET: Transverse fetal chest position, size, rhythm; four-chamber view (apical + subcostal); long-axis view of the aorta; long-axis view of the pulmonary artery (or short-axis Ultrasound in Obstetrics and Gynecology 375

7 view of the great vessels or three-vessel view); short-axis view of the great vessels; aortic arch; ductal arch (or short-axis view of the great vessels). UG: The four-chamber view is the basis of cardiac examination in a screening setting. Besides the cardiac structures the size and the position of the heart and the heart rhythm should be checked. In the future, however, the examiner should also be trained to visualize the ventricular outflow tracts with the origin and course of the great arteries and incorporate these views in the screening examination. Especially the continuity of the interventricular septum to the anterior wall of the ascending aorta should be imaged in a longer axis view of the left ventricle and the size and the course of the great arteries with the aortic arch should be visualized in the three-vessel view (arterial duct, aortic arch and superior caval vein) by a transverse plane in the upper thorax and/or by the classic short-axis view of the aorta at the cardiac base ( circle and sausage view). JH: Four-chamber and outflow tracts. GP: Evaluation of the great vessels improves the detection rate of the four-chamber view, but realistically remains beyond the capability of the vast majority of sonographers performing fetal sonographic screening. In my view, a policy of careful standardization of the elements that constitute the four-chamber view (position of the heart inside the chest, correlation with abdominal organs, cardiac axis, size of the chambers, integrity of septa, use of cine-loop and/or color Doppler to demonstrate the opening movement of the atrioventricular valves) could be more productive than the compulsive use of great-vessel views by sonographers who do not clearly understand the rather complex tomographic anatomy of the ventriculo-arterial connections. JW: The screening examination should include the following cardiac views: four-chamber view, short-axis and long-axis views. SY: My recommendation for optimal cardiac screening is for five short-axis views of the fetal upper abdomen and mediastinum. These are described in detail in the Editorial of this issue of the Journal 6. Question 7: Who should do the screening: obstetrician, specialist perinatologist, radiologist, sonographer or technician? LA: If properly trained and supported by a specialist fetal cardiologist, any of the above are perfectly capable of routine screening. None of the above are competent without proper focused training and practise. BB: All of the above personnel should be involved in the screening of the fetal heart. Whoever has taken the time and made the effort to become proficient at the evaluation of the fetal heart should be involved in performing the exam. This may be an obstetrician, a perinatologist or a radiologist and certainly all sonographers should become skilled in this technique. The evaluation of the fetal heart is a team effort and only the degree of training/experience separates those that become successful at doing it from those that don t. JAC: The RADIUS study 7 showed that fetal imaging specialists can achieve better sensitivity than non-specialists. But that imaging specialist might be obstetrician, perinatologist or radiologist. Practically speaking in the US most are specialist perinatologists and radiologists. I believe that the differences relate in part to volume of fetal anomalies seen as well as the interest of the sonologist and sonographer performing the sonograms. JSC: It does not matter who the professional is. What matters is that he or she is confident in assessing the fetal heart. RC: Screening should be done by trained persons understanding the whole fetal anatomy on ultrasound and performing many scans a week. All cited persons can perform it, under these conditions. Occasional examinations cannot lead to an improvement in quality. However, it has to be borne in mind that by definition screening should be an easy and cheap method with a high sensitivity. Therefore it is not optimal to let very experienced persons, like specialist perinatologists and radiologists (or pediatric cardiologists), do screening on a low-risk population. SE-N + ET: Depends on how the department is organized. Sonographers or other educated personnel such as specialist midwives. UG: Who should do the cardiac screening is mainly dependent on the organizing structures for prenatal screening in each country. It seems to be reasonable that the cardiac screen as part of the general malformation screening by ultrasound should be performed by one highly experienced examiner. Improvement in the level of expertise by permanent training and close liaison between the screening examiner and the specialized referral center are mandatory to exhaust the possible diagnostic potential of probably the most important examination in all fetal and human life. GP: Great variations in the organization of health care exist in different countries. It is unrealistic to employ highly specialized and skilled professionals such as pediatric cardiologists and perinatologists for universal screening of pregnant patients. Screening should be performed by personnel expert in obstetric scanning. JW: The screening should preferably be done by a specialist perinatologist, to ensure high quality morphological and, if necessary, functional assessment of the entire fetus. SY: The initial screening can be performed by any practitioner who has received the appropriate training, be it a sonographer, radiologist, obstetrician or other specialist. The training and expertise, not the degrees held by the examiner, are of paramount importance. Question 8: Is a pediatric cardiologist essential for prenatal diagnosis and counseling? LA: Yes. The results of treatment for MCHD are changing all the time and the only people in a position to be abreast of this constantly are pediatric cardiologists. However, they also should not dabble in fetal diagnosis or counseling but be experienced in and dedicated to the speciality and trained appropriately. BB: No, I do not believe that a pediatric cardiologist is required to make the prenatal diagnosis of MCHD. As seen above in my answer to question 7, the practitioner who has gained expertise in the evaluation of the fetal heart is the 376 Ultrasound in Obstetrics and Gynecology

8 person most essential, whether it be a cardiologist, perinatologist, obstetrician or radiologist. In my own practice, I do refer the patients to the pediatric cardiology team after I have made a specific diagnosis of MCHD, especially for more in-depth counseling about the postnatal outcome and for additional expertise on complex lesions. I believe in the team approach to diagnosing MCHD and patients benefit greatly from cardiac consultation even if the diagnosis is already known. Also, these same pediatric cardiologists will be caring for the newborn after birth, thus it is beneficial for the patient to meet these physicians and establish a relationship early. JAC: Yes. The role however, will vary with the skill and interest of the cardiologist. I believe some need to scan themselves to gain full appreciation of the lesion, while others are less comfortable scanning and prefer to review images obtained by the fetal imaging expert. The knowledge base and experience of the pediatric cardiologist, however, make them critical to complete counseling of the family. JSC: Unquestionably YES! It is also important that the pediatric cardiologist has experience in diagnosing CHD in the fetus. For counseling, no other specialist will (or could) have had the experience of diagnosing and treating neonates, children, adolescents and young adults with CHD. RC: Only for rare findings could the cardiologist be essential for diagnosis. Most anomalies are, however, already known and could be diagnosed accurately by an experienced non-cardiologist. On the other hand counseling should be performed in cooperation with a pediatric cardiologist. He or she is best informed on the actual data of shortand long-term outcome and knows the results at his/ her center and not only the data from the literature. In fetuses with (isolated) cardiac defects a second opinion should be obtained in all cases. SE-N + ET: In our center the scanning is done by an obstetrician in close cooperation with pediatric cardiologists when a CHD has been diagnosed. Who does the scan is irrelevant. The main thing is to have cooperation with a pediatric cardiologist for interpretation of the diagnosis, prognosis and general patient counseling. UG: Adequate prenatal counseling of parents after in utero diagnosis is dependent on many factors, in particular on the accuracy and security of diagnosis of the type of CHD, the association of extracardiac malformations, the natural history in fetal and postpartum life, the treatment options and the gestational age at diagnosis. This demands close cooperation between the specialist for fetal diagnosis and medicine and the local pediatric cardiologist. Therefore, after exclusion of severe or even lethal extracardiac anomalies, counseling of the parents of fetuses with diagnosed cardiac defects should always be performed together with the pediatric cardiologist who is familiar with the modern treatment options for special MCHDs. On the other hand, if a cardiac malformation is suspected the detailed echocardiographic fetal examination for achieving a precise cardiac diagnosis should be performed by the person who has the most experience in fetal heart examination. Depending on the different organizational structures in different countries it may be the specialized pediatric cardiologist or the specialist on fetal ultrasound diagnosis who is familiar with both cardiac and extracardiac ultrasound examination. In particular, early echocardiographic examination by transvaginal scanning appears to be unfamiliar to most pediatric cardiologists. Therefore, I think that the continuous training in fetal scanning and also the increasing expertise in cardiac ultrasound favored by the high quality of modern ultrasound equipment will lead to the situation where fetal echocardiography will be performed more and more by highly specialized obstetric ultrasound examiners. JH: Yes. Bergella et al. 8 showed that the cardiologist s detection rate and accuracy were better. Working together is the key. GP: Anyone who is adequately trained in obstetric ultrasound can make an accurate diagnosis of even complex fetal cardiac malformation, independently from their background. The collaboration with a pediatric cardiologist can be useful at times but certainly it is not essential. Counseling is a different matter, and it would certainly be advisable to involve both a perinatologist and a pediatric cardiologist. JW: I believe that a pediatric cardiologist can be helpful in establishing prenatal diagnosis and is essential for adequate counseling. SY: Yes, the input of a pediatric cardiologist in the counseling of prospective parents of a fetus diagnosed with MCHD is essential. Their expertise is invaluable on the subjects of prognosis of various lesions, the surgical or other interventions that will be required over the years, and the child s expected quality of life; all these are the province of the pediatric cardiologist. Ideally, parents should receive counseling from a team of specialists, including the mother s physician, the diagnosing examiner, a geneticist, and a pediatric cardiologist. The diagnosis of the lesion must be done by a physician who specializes in prenatal diagnosis of fetal heart disease, whether a pediatric cardiologist, specialist perinatologist, obstetrician, or other senior consultant. Question 9: What are the antenatal predictive criteria for MCHD at 20 weeks gestation that would justify counseling towards a good postnatal outcome? LA: As a general rule, heart disease in which a two-ventricle repair seems possible is much more hopeful in terms of longterm prognosis than the type of disease which has to go a oneventricle (Fontan repair) route. A Fontan repair is known to be palliative in childhood and problems with this type of circulation are likely to arise in early adult life. In addition, lesions which can be repaired without the use of prostheses, homograft or conduit, which of necessity require replacement with growth during childhood, are associated with less morbidity. BB: The MCHD lesions with the best prognoses are those that can result in two-ventricle physiology after intervention. Defects where the expectation after surgery is for single-ventricle physiology have the worst outcomes. I tend to predict prognosis based on this premise. JAC: Two ventricles, normal karyotype and no other significant fetal anomalies. Intermediate prognosis: need for major great artery reconstruction but still two usable ventricles. Poor prognosis: only one operable ventricle. Ultrasound in Obstetrics and Gynecology 377